Fuzzy neural network (FNN)-based model analysis techniques for enhancing the development evaluation for switched reluctance motor (SRM) drives

Abstract

in this paper, an analytical methodology utilizing fuzzy neural network (FNN) that improves evaluation capabilities for switched reluctance motor (SRM) drives is presented. Firstly, a FNN is applied to establish the mapping relation among the driving current, rotor angle, and equivalent inductance by the information obtained from the SRM characteristics measurements. Then two FNNs, which own the competence to do partial derivative computation, are constructed herein to complete this FNN-based analysis model. The computation results under different speed commands and load conditions respectively verify the proposed analysis opinions.